Modeling and simulation of lightweight lattice structures for impact energy absorption

Metal lattice structures possess a wide range of applications in the mechanical field, mainly because of their high functionality and their lightweight. Additive Manufacturing (AM) offers an incredible design freedom for topologically complex components, that before the advent of this technology were manufactured with many difficulties and expenses: world of lattices made a huge step forward thanks to AM. Different engineering disciplines employ lattices for designing components, using latest technologies. One of the most important application nowadays for these materials, especially when it comes to aerospace, biomedical and automotive sectors, is as energy absorbers: this thesis wants to be a contribution in this direction. In particular, the main aim is to establish design guidelines for truss-based lattice, when contributions of load directed reinforcements and cells combination are present; multi-morphology design will be considered as well. A specific set of lattice cells, selected among the most promising ones for energy absorption, has been tested using FEA (Ansys). As a starting point for investigations, static simulations have been performed using a bilinear model for the material used (that is AlSi10Mg, one of the most used both for its lightweight and for the fact that is easily processable using AM). Bilinear model is ideal in this case because it allows with very good approximation the study of the plastic domain of the material. Most of the energy absorption characteristics of lattices lies in fact in their plastic filed and it ends with densification; simulations are performed taking account of this. Results from static simulations are therefore used for a confrontation with results from high speed impact simulations, whose model is basically inspired by classical crashworthiness and impact tests. These simulations, performed using the Ansys explicit solver, offer answers about densification and its effects (impossible to catch with static simulations) and energy parameters needed to rank and to analyze the selected lattices. One of the main objectives of these analyses, beside just mentioned results, is to check if static analyses are able to predict dynamic response. When it comes to analyses of large samples or components, modeling methods contemplating generation of the whole lattice geometry can present extremely long solving process times. For sake of time but also for sake of the nature of the simulations themselves (that should allow time and money saving), a homogenization process has been used in order to make multi-morphology analyses more affordable: one single cell can be in fact modeled as one, or more, element. The developed method allows the creation of a medium material that brings with him information about the lattice elastoplastic behavior, because, as said, the plastic domain makes the difference in this situation. Moreover, errors and deviations from the original models have been evaluated, in order to be aware of the method limits. Therefore, samples presenting configurations inspired by classical Reuss, Voigt and lamination theories are generated and tested under high speed crashing conditions using explicit solver. These simulations offer results about how mixing lattices with specific rules can improve energy absorption. Results obtained are therefore used to establish construction guidelines and a methodology for lattice simulations that can be valid for energy absorption analysis.